2 ms·
I'm a Prolog newbie myself, but an experienced ML practitioner. Doing "high-level" model-building in a Prolog-like DSL might be interesting, but lower-level nu
by nerdponx 6y ago
I'm a Prolog newbie myself, but an experienced ML practitioner.
Doing "high-level" model-building in a Prolog-like DSL might be interesting, but lower-level numerical code seems like a chore.
The real "value" of Prolog (in my novice estimation) is querying data (i.e. Datalog), transforming between data formats, and building "logic-driven" applications like command-line tools, API clients, etc.
- YeGoblynQueenne 6y ago>> I'm a Prolog newbie myself, but an experienced ML practitioner. How experienced? Inductive Logic Programming has been a thing since 1991 :-)
- nerdponx 6y agoLet's go with "ML as in applied statistics", not the traditional stuff.
- YeGoblynQueenne 6y agoFair enough, that's the common interpretation.
- nerdponx 6y agoI should have known to be careful with that term in a Prolog thread!
- YeGoblynQueenne 6y agoNah, I'm just a compulsive pedant :P